Applying Six Sigma in Higher Education Quality Improvement · Six Sigma Methodology: Statistically Six Sigma quality defines limiting the number of defects to 3.4 (parts per million

Paper ID #8594

Applying Six Sigma in Higher Education Quality Improvement

Dr. Quamrul H. Mazumder, University of Michigan, Flint

Dr. Quamrul Mazumder has been conducting research in the areas of metacognition, teaching and learningstyles, motivation and engagements. As a Fulbright scholar, he was involved in higher education qualityimprovement initiatives in Bangladesh. He published a book titled ”Academic Enhancement in HigherEducation”.

c©American Society for Engineering Education, 2014

Page 24.191.1

Applying Six Sigma in Higher Education Quality Improvement

Abstract

Quality in higher education became an important issue due to ever increasing demand by

stakeholders and competitive environment. Although six sigma has been successfully used in

product and service improvement in the business environment, the concept has not been adapted

in higher education. To improve understanding of how six sigma can be used for higher

education process improvement toward achievement of quality, a number of models are

presented. Six sigma principles such as process improvement, reducing waste and continuous

improvement aligns closely with the mission of higher education institutions and accreditation

agencies. Using six sigma tools such as statistical process control, lean manufacturing, failure

mode and effects analysis can help in the development of sustainable higher quality educational

process. A process map with SIPOC (supplier, input, process, output and control), cause and

effect analysis, FMEA (failure mode and effects analysis) for higher education was developed

and presented. These tools can be used by higher education institutions to better understand the

higher education process and how it can be improved to meet the desired quality goals.

Introduction

The concept of Six Sigma was introduced by Motorola in the 1980s to improve their products

and maintain quality. The core of Six Sigma lies in the continuous improvement process using

the DMAIC (Define, Measure, Analyze, Improve, and Control) method [9]. It has since then been

adopted by many other companies to achieve their respective goals both in production of goods

and in rendering services. Due to the success of this method, academic institutions attempted to

adapt six sigma methodologies to improve the quality of education and services. These concepts

have great potential for improving process efficiency and quality of higher education. The

improvements can be enhanced by integrating other similar concepts such as lean manufacturing

and SPS (statistical process control).

Lean manufacturing was originated as “a philosophy of continuously simplifying processes and

eliminating waste”[16]. By streamlining the processes, cycle times for data collection and analysis

can be reduced in academic environment due to time constraints faced by students and faculty.

The statistical process control (SPC) method uses control charts to analyze variations in a

process with predetermined upper and lower control limits (UCL, LCL). Two types of variations

are common in any process and are described as follows: (1) random variations, which are the

only variations present if the process is in statistical control, and (2) assignable variations, which

indicate a departure or deviation from statistical control. The purpose of a control chart is to

identify when the process is out of control, thus signaling the need for remedial action. A control

chart is a graphical technique in which statistical results are computed from measured values of a

certain process characteristic are plotted over time to determine if the process remains in

statistical control. Statistical process control charts and run charts are helpful tools for large

amounts of outputs such as in manufacturing processes or when dealing with a large student

body in a university [12].

Page 24.191.2

Literature Review:

According to Freeman, there is an increasing need to improve the quality of higher education

because education is becoming a global entity facing challenges with resource constraints [3].

Unlike other organizations, higher education has several stakeholders such as students, parents,

future employers and society [7]. Zhang proposed eight important questions to ask regarding a Six

Sigma research program. Of these eight, the most relevant to higher education are: “How can the

effectiveness of a Six Sigma program be validated?” “How should Six Sigma be customized for

different organizational contexts?”, “What is the most effective organizational structure for a Six

Sigma program?”, and “How do leadership development and human resource practices relate to

Six Sigma program?”[19]. The answers to these questions center on empirical validation of

effectiveness and customization of the program, separating the Six Sigma program from Quality

Control

Adaptation of six sigma approaches in higher education requires careful consideration of

differences in stakeholders’ requirements and expectations. Unlike business environment,

higher education may be perceived by some as non-profit to serve the greater intellectual and

societal needs. Decisions in higher education are not always data driven and the need for data is

underestimated. An example of a process improvement involves recording scores on the

accounting section of the Educational Testing Service standardized test. Additional data such as

faculty assignments, textbooks, course design, teaching methods, and course order were

collected. To improve average test scores from 42.4% to 46.5%, the input variables were

altered. These changes to the program design resulted in an actual increase to 47.3%, above the

desired goal[6].

In the study by Razaki & Aydin, different process improvement methods from the business

world are analyzed for their usefulness in the academic world. Four different methods were

analyzed, including Total Quality Management (TQM), Six Sigma, Business Process

Reengineering (BPR), and Lean Manufacturing. “TQM was highly suited to improving the

departmental processes to effect a transition to excellence, Lean Six Sigma provided a few but

highly effective methods for departmental improvement.” The use of Lean Six Sigma was

revealed from their analysis of the Kukreja study. It was noticed that the data collection cycle

was too long and a great deal of time was necessary to complete the project. Since most students

are only enrolled for four years, this did not work well with this required timespan. They propose

mixing the appropriate parts of Six Sigma and Lean Manufacturing to make the process more

appropriate for the relatively short time available to collect data on individuals. This method uses

statistical tools of moderate complexity, with a short cycle time and a focus on elimination of

waste [12].

Higher education process can be viewed to be similar to a manufacturing process. In a

manufacturing process, raw materials are processed through a series of steps to produce finished

products. Similarly, the higher education institutions produce intellectual graduates from

incoming students through a series of steps. In higher education, quality depends on several

factors such as curriculum, course content, incoming students, teachers, pedagogy, and

assessment methods. Since one of the focuses of Lean Manufacturing is reducing waste, it is

important to define waste in the higher education system of processes. Examples of educational

Page 24.191.3

waste include, “teaching topics already taught in other courses, excessive review of prerequisite

materials, unnecessary and redundant introductions, spoon-feeding, teaching obsolete topics, and

waiting for unprepared students to catch up” [16]. In order to produce a high quality graduate,

efforts to minimize wastes must be undertaken throughout the process with careful consideration

of stakeholders’ views.

Statistical process control can be a useful tool in the academic environment as the institutional

analysis involves a large amount of data such as enrollment trends, graduation rate, retention

rates, etc. As every process has an expected degree of variation, it is necessary to determine

what constitutes ‘normal’ variation so that it can be predicted. The more the variation of a

process can be minimized or controlled, the more accurately the process results can be

predicted.” When the process is under appropriate control, the produced variations will be

consistent and within the accepted range. The method of SPC can be challenging to apply outside

a manufacturing environment, such as a service industry like higher education. In situations

where performance parameters are not taken from tangible, measurable products more work is

needed. In a study by Roes & Dorr of SPC implementation in the service industry, the key

characteristics for process control were defined as the degree to which the service to the

customer is indeed intangible, the intensity of involvement of employees in the interaction, and

the extent of customer influence on the service provided [13]. For academic environment, the

customer would be a future employer, employees would be university faculty, and the service

would be the provided education. The SPC approach can be used to improve course instruction,

using the following steps:

1. Identify the process to control

2. Determine quality characteristic to monitor

3. Choose the appropriate control chart based on

a. Type of data

b. Sample size

c. Frequency

4. Perform process improvement using SPC tools

5. Implement continuous quality improvement on process [10].

Quality, with respect to higher education has several challenges such as endurance, conformance

to requirements, continuous improvement and value added [2]. The process variability not only

exists within the students, but within professors as well. For example, grading by professors may

be different and the instructional methods may also have variations. In a study by Knight,

professors graded unnamed assignments and then re-graded these assignments weeks later to

observe the difference in grades received. These grades were then subjected to statistical

analysis, finding the average range for each professor. These were then averaged with each other

and used to find an upper control limit for the ranges themselves. In future grading, if grades

exceeded this range, the assignments would then be re-evaluated [5]

The application of Six Sigma DMAIC methodology to improve quality in engineering

educational had been successful in improving the quality consciousness with students and the

management of institution [11]. The Six Sigma method can also be applied within the course to

continuously improve its quality. The Statistics department of Florida State University, engaged

students in seven different projects throughout a course. The first project involved the students

listing two contributions they would like to make to their careers. The next five projects followed

Page 24.191.4

the DMAIC process, and the final project requires a report on the overall process. In each

project, the students applied the DMAIC principles toward achieving their goal, learning the

language and function of Six Sigma as they progress [18]. By applying DMAIC, students were

able to achieve their goals and familiarize themselves with the system.

The problems associated with change management is challenging in higher education due to the

nature of the environment that promotes academic freedom. Academicians have been

accustomed with this environment and have individual views towards different issues as well as

departmental politics and inter-departmental acrimony that increase complexities associated with

any change in the process. “It is estimated that 70% of organizational change initiatives fail

completely. Of the ones deemed successful as many as 75% of these fail to achieve their

intended result.” Individuals do not always get along in an organization, and, when the success of

a program is dependent on collaboration, noncooperation can be a hindrance to achieving an

organizational change. Given all these problems, there is a question as to whether there truly is a

“best practice” for such change. “It appears that many popular management practices labeled as

best practices (such as Total Quality Management, Six Sigma, and Lean) are based on anecdotal

evidence rather than empirical data.” This perception may be due in part to the fact that “the

terms ‘organizational change,’ ‘change management,’ and ‘best practice’ appear to be used in a

variety of perspectives and research applications but the search for affinity patterns have not

resulted in any stable conclusions”[4]. As with many aspects of management, it would appear that

flexibilities must be exercised to implement changes appropriate to the environment. Apart from

the students, teachers and the management involved, the infrastructure and educational resources

that students access also proves to be vital in achieving a higher quality education [14].

Six Sigma Methodology:

Statistically Six Sigma quality defines limiting the number of defects to 3.4 (parts per million

PPM). The term Six Sigma refers to the six standard deviations away from the mean in a normal

distribution or bell shaped curve. It uses the measurement of factors in a process and works on

improving the output based on continuously improving the system and its processes. The defects

in a Six Sigma process are the total area to the right and left of +6σ and -6σ respectively as

shown in Figure 1.

Page 24.191.5

Figure 1: A normal distribution curve with six sigma (σ = 0 at mean)

Among different approaches used towards achieving six sigma level of quality, the DMAIC and

the focus is on continuous improvement lies in the heart of six sigma process. DMAIC is an

abbreviation for Define, Measure, Analyze, Improve and Control. The following section of the

paper attempts to demonstrate how DMAIC methodology can be used to continuously improve

the quality in higher education.

Define Phase:

In the design phase, the goals and the parameters must be clearly identified and defined. Six

Sigma methodology can be effectively used in higher education institutions [1]. The first step to

understanding the process is to develop a process map for higher education and then construct a

cause-effect diagram to evaluate the effect of input variables on output. A process map for higher

education is presented in figure 2 and compared to a manufacturing process as shown in figure 3.

The potential suppliers of higher education are educational institutions such as high schools,

community colleges or universities. The input consists of new first year students, transfer

students, K-12 teachers, and high school graduates. The Process involves a sequence of steps

from which a student takes various course over a period of time and graduates. The customers

consist of employers, graduate schools, society, and others, as some students may be self-

employed.

Page 24.191.6

Figure 2: Six Sigma Process (SIPOC) in Higher Education

Figure 3: Process flow in a conventional manufacturing process

Measure Phase:

In the measure phase, all measurements related to the process are calculated. Although a number

of different measurement tools can be used in this phase, an example of SPC is presented in this

paper. Among different factors affecting quality of education process and student performance,

the important ones may be GPA, professors’ performance, number of students in each class,

course materials and course order. The factors used to measure student success are student

retention rate, graduation rate, and percent employed in the field related to academic degree

immediately after graduation as presented in Figure 4. These variables can be analyzed using

SPC to identify which input or inputs have the greatest effect on the outputs. Some of the inputs

do have dependencies on each other and this will be analyzed as well to ensure the accuracy of

the analysis.

Figure 4: Output Controls and its dependency on Inputs

Page 24.191.7

Both quantitative and qualitative control charts have been developed to monitor the performance

of individual student and the institution. The two quantitative charts are the Individual/Moving

Range chart (IX/mR) and the Average/Range chart (X̅/R). To monitor an individual’s

performance IX/mR chart was developed using the following steps.

1. Gather the data. (Verify data validity by considering the collection method.)

2. Calculate the moving ranges (difference between each successive data point).

3. Plot the data in time ordered series (Individuals [IX] chart)

4. Plot the moving ranges in time ordered series on the moving range (mR) chart.

5. Calculate the following formulas provided on the following pages:

a. Average of all the moving ranges (mR̅̅̅̅̅) b. Estimate of the sigma/standard deviation (mR̅̅̅̅̅/d2)

c. Average of all the data points (X̅)

6. Plot the lines representing the averages, LCL’s, and UCL’s on the IX and mR charts.

Table 1: Courses and GPA in the class with moving range (mR)

Class Level Course GPA Received mR

Freshman EGR102 4

Freshman EGR280 3.3 0.7

Freshman EGR230 3.7 0.4

Freshman EGR260 3.3 0.4

Freshman EGR165 4 0.7

Sophomore EGR310 3 1

Sophomore EGR350 3 0

Sophomore EGR353 2 1

Sophomore EGR330 3.3 1.4

Sophomore EGR356 3.7 0.4

Junior EGR370 2.7 1

Junior EGR315 2.7 0

Junior EGR321 2 0.7

Junior EGR392 1.3 0.7

Junior EGR380 2.7 1.4

Senior EGR399 3.3 0.6

Senior EGR432 3.7 0.4

Senior EGR410 3 0.7

Senior EGR465 3.7 0.7

The current study for an individual student was based on the grade point average (GPA) in the

courses related to their major. The students’ academic progress was considered as a single

process for application of SPC recognizing variations in courses, professors, and levels. For

example, a student in the Mechanical Engineering program requires five prerequisite courses,

eight core courses, and seven elective courses with a total of 20 engineering courses. Table 1

shows the moving range chart for GPA in the engineering courses. The upper control limit Page 24.191.8

(UCL) and lower control limit (LCL) of moving range and individual control chart is presented

in Table 2.

Table 2: UCL and LCL of Moving range and Individual Control Chart

Moving Range (mR) Chart Data

Average mR 0.673684

Estimate of Sigma 0.597232

UCL (mR) 2.200963

LCL (mR) 0

Individuals (IX) Chart data

Average GPA 3.07

UCL (IX) 4.861713

LCL (IX) 1.278287

Figure 5: Average GPA of students in each course

A control chart for students’ GPA in different engineering courses is presented in Figure 5. The

UCL of 4.0 represents the maximum attainable GPA and the LCL of 2.0 represents the minimum

required GPA required by the university to be in good standing. Students with less than a 2.0

GPA are placed on academic probation and may be terminated if they fail to improve their GPA.

It can be observed from the control chart that the process is not in control for two courses. The

average GPA for EGR 353 (Thermodynamics) is at the lower limit of control chart and the

average GPA for EGR 392 is lower than the lower control limit. This clearly identifies

improvement needs in these two courses, as the success rate of students in terms of GPA is less

than expected. Figure 6 shows the moving range showing any significant difference between

two successive control points. All the points in the moving range chart are within the control

1

1.5

2

2.5

3

3.5

4

4.5

Av

era

ge

GP

A

Name of course

Average GPA

UCL

LCL

Mean

Page 24.191.9

limits and hence there is no control point with significant difference compared to its successive

control point.

Figure 6: Moving Range (mR) across each course

Figure 7: Normal Distribution (Bell Curve) of student’s grade (GPA)

Figure 7 shows the normal distribution curve for the student’s grade. The right side of the curve

has a maximum value of 4 with the minimum value as zero. The average value of GPA (µ) from

the current set of data was 3.07 with a standard deviation (σ) of 0.7027. The students with less

than a 2.0 GPA may be considered as the defects in the system as shown in the area of normal

distribution curve to the left of X = 2. The area to the left of X = 2 was calculated as 0.06392

which means that approximately 6.4% of the students received GPA of less than 2. Therefore,

the defects per million is 63,920 that meets 3σ level of quality in the process. To achieve six

sigma level of quality the value must be reduced significantly.

0

0.5

1

1.5

2

2.5M

ov

ing

Ra

ng

e (m

R)

Name of course

Average mR

UCL

LCL

Mean

Page 24.191.10

Analysis Phase:

After the development of the process map, it is important to identify the causes for poor quality

in higher education. A cause and effect or fishbone diagram is a widely used approach to

identifying the root causes and their effects. The sources of poor quality were identified as

curriculum, teachers, students, assessment, and the academic and social environment. The

possible causes from each of these sources have been schematically shown in figure 8. The

fishbone diagram displays the root causes from six different sources that contribute to poor

quality of education. Identification of these sources can help in making changes to improve

quality of education.

Figure 8: Cause and Effect Diagram of Quality of Higher Education

Improvement Phase:

In the improvement phase, the causes for failure or poor quality must be identified with a

solution that will reduce defects in the process. A failure mode and effect analysis or FMEA can

be used to improve the process. These quality tools could be very well used for the improvement

of organizations and institutions [17]. A step-by-step procedure is used to identify all possible

causes of failure and their corresponding effects with recommended corrective actions to avoid

the failure modes. Quality needs to be properly assessed with respect to students, teachers,

Page 24.191.11

departments and Institutions, which makes curriculum [15]. A FMEA was developed to address

the above factors as shown in Table 3.

Table 3: A Failure Mode and Effect Analysis of Higher Education Process

Page 24.191.12

Control Phase:

The control phase requires institutionalization of the improvement results obtained from the Six

Sigma process for sustainability. The key to success in achieving quality is to standardize the

improvement process and fostering a six sigma or continuous improvement process in the

organizational culture. The results of the new standardizations or procedures can be further

improved using different six sigma tools and procedures with a goal of reducing variation or

defect in the process. Control charts are an effective way of statistically keeping a track of

performance and using the data for continuous improvement in Six Sigma methodology [8].

Summary

A number of six sigma models have been developed and presented to improve quality in higher

education. The key inputs and output variables were identified in the define phase of DMAIC

process. The input and output variables were measured by collecting the data over time. The

analysis phase used SPC to identify the variables outside the control limits. After identification

of the variable that lies outside the control limits, appropriate corrective actions can be

implemented for process improvement. This phase is considered important in academic

environment, as it is critical to student success and quality improvement. In the control phase, the

input and output variables require continuous monitoring to ensure sustainable process.

Conclusion

The higher education process showed a three sigma (3σ) level quality that requires significant

improvement to achieve six sigma (6σ) level. The primary objective of higher education is

student success through higher quality education where failure of any student may be considered

as a defect in the process. Due to variability in the process such as different type of instruction

by different professors, a variation of quality exists. Variations of quality may be due to lack of

understanding of how students learn and adapting to different learning styles of students. After

identification of the issues and defining the problems, a solution can be developed using six

sigma approaches and models presented in this paper. A control chart can be used with UCL and

LCL along with a continuous improvement plan to improve the higher education process. This

will result in higher quality and sustainable process in the institution with higher levels of student

satisfaction and success rates such as graduation and retention rates. The information and tools

provided in this paper is an attempt to shed some lights on how different quality improvement

models can be used in higher education.

References

[1] Antony, J., Krishan, N., Cullen, M., & Kumar, M. (2012). Lean six sigma for higher education institutions (heis):

Challenges, barriers, success factors, tools/techniques. International Journal of Productivity and Performance

Management, 61(8), 940-948.

[2] Dew, J. (2009). Quality issues in higher education. The Journal for Quality and Participation, 32(1), 4-9.

Page 24.191.13

[3] Freeman, R. (1993). Quality assurance in training and education. (p. 176). London: Koogan Page.

[4] Hallencreutz, J., & Turner, D. (2011). Exploring organizational change best practice: are there any clear-cut

models and definitions. International Journal of Quality and Service Sciences, 3(1), 60-68.

[5] Knight, J. E., Allen, S., & Tracy, D. L. (2010). Using six sigma methods to evaluate the reliability of a teaching

assessment rubric. The Journal for American Academy of Research Cambridge, 15(1), 1-6.

[6] Kukreja, A., Ricks, J. M., & Meyer, J. A. (2009). Using Six Sigma for performance improvement in business

curriculum: A case study. Performance Improvement, 48(2), 9-25.

[7] Madu, C. N., & Kuei, C. H. (1993). Dimensions of quality teaching in higher institutions. Total Quality

Management, 4(3), 325-338.

[8] Maleyeff, J., & Kaminsky, F. (2002). Six sigma and introductory statistics education. Education Training, 44(2),

82-89.

[9] Mitra, A. (2004). Six sigma education: a critical role for academia. The TQM Magazine, 16(4), 293-302.

[10] Perry, L. (2004). Instructional effectiveness: A real-time feedback approach using statistical process control

(spc). Proceedings of the 2004 American society for engineering education annual conference & exposition, Utah,

USA.

[11] Prasad, K. D., Subbaiah, K. V., & Padmavathi, G. (2012). Application of Six Sigma Methodology in an

Engineering Educational Institution. Int. J. Emerg. Sci, 2(2), 222-237.

[12] Razaki, K. A., & Aydin, S. (2011). The Feasibility of Using Business Process Improvement Approaches to

Improve an Academic Department. Journal of Higher Education Theory and Practice, 11(2), 19-32.

[13] Roes, K. C., & Dorr, D. (1997). Implementing statistical process control in service processes. International

Journal of Quality Science, 2(3), 149-166.

[14] Sasikala, S., & Vincent, G. S. (2010). Infrastructure and learning resources in higher educational institutions

(HEIS) using six sigma quality strategy. Library Progress (International), 30(1), 97-109.

[15] Stark, J. S., & Lowther, M. A. (1980). Measuring higher education quality. Research in Higher

Education, 13(3), 283-287.

[16] Tatikonda, L. (2007). Applying Lean Principles to Design Teach, and Assess Courses. Management Accounting

Quarterly, 8(3), 27-38.

[17] Weinstein, L. B., Petrick, J., Castellano, J., & Vokurka, R. J. (2008). Integrating Six Sigma concepts in an

MBA quality management class. Journal of Education for Business, 83(4), 233-238.

[18] Zahn, D. (2003). What influence is the six sigma movement having in universities? what influence should it be

having?. ASQ Six Sigma Forum, 3(1), Retrieved from

http://asq.org/pub/sixsigma/past/vol3_issue1/youropinion.html

[19] Zhang, W., Hill, A. V., & Gilbreath, G. H. (2011). A research agenda for Six Sigma Research. Quality

management journal, 18(1), 39-53.

Page 24.191.14